Method for the continuous measurement of the glucose concentration in blood undergoing pulsational flow, with the steps:—determination of a value for the glucose concentration for a first measurement cycle, and—repetition of the determination of this value in subsequent measurement cycles, where there is multiple detection, within each measurement cycle, of the transmittance and/or scattering power of the blood for at least two incident NIR wavelengths, calculation of an indicator value depending on the blood glucose concentration, and ascertaining the blood glucose concentration by comparing the indicator value with a previously determined calibration table, determination of the blood temperature during the detection of the transmittance and/or scattering power,—continuous measurement of the pulse duration of the pulsational blood flow, where the duration of the measurement cycle is arranged to keep in step as integral multiple of the pulse duration, where the first of the at least two NIR wavelengths is selected from the wavelength range 1560-1630 nm, and the second of the at least two NIR wavelengths is selected from the wavelength range 790-815 nm, and the ratio of the transmittance and/or scattering power of the at least two wavelengths is calculated, this ratio serving in relation to the blood temperature as indicator value for reading off the blood glucose concentration from the calibration table.
